The continual demand for enhanced integrated circuit performance has resulted in, among other things, a dramatic reduction of semiconductor device geometries, and continual efforts to optimize the performance of every substructure within any semiconductor device. A number of improvements and innovations in fabrication processes, material composition, and layout of the active circuit levels of a semiconductor device have resulted in very high-density circuit designs. Increasingly dense circuit design has not only improved a number of performance characteristics, it has also increased the importance of, and attention to, semiconductor material properties and behaviors.
The increased packing density of the integrated circuit generates numerous challenges to the semiconductor manufacturing process. Nearly every device must be smaller without degrading operational performance of the integrated circuitry. High packing density, low heat generation, and low power consumption, with good reliability must be maintained without any functional degradation. Increased packing density of integrated circuits is usually accompanied by smaller feature size.
As integrated circuits become denser, the widths of metal structures interconnecting transistors and other devices within an integrated circuit are reduced. As the width of metal interconnects decrease, their resistance increases. As a result, semiconductor manufacturers seek to create smaller and faster devices by using newer, alternative materials (e.g., copper, tungsten) to form metallic interconnect structures, instead of more traditional semiconductor materials (e.g., aluminum).
The introduction of alternative materials, such as copper, has successfully addressed certain performance concerns—so much so that those materials are now used extensively throughout many manufacturing processes. Performance advantages provided by such materials, however, have not come without cost. Certain physical and behavioral properties of these alternative materials can cause of number of problems during the manufacture, testing or operation of a semiconductor. Copper, for example, is known to be an extremely active and atomically mobile material. In many semiconductor devices, copper can and will diffusively migrate from a metallic structure to a collateral, non-metallic structure unless certain measures are taken to limit such diffusion. Copper contamination of non-cupric semiconductor materials can significantly degrade device parametric performance, or cause complete device failure altogether. This naturally results in a number of yield and reliability problems.
Intra-device copper diffusion is not the only potential source of copper contamination problems. In a number of instances, non-cupric semiconductor materials—especially front-end materials (e.g., silicon compounds, oxides) can become contaminated with copper—generally a back-end metallization material—at a variety of stages during the fabrication process. Cross-contamination can occur in or at a number of processing systems or stations (e.g., device inspection, cleaning chambers), since such systems must be used to process multiple devices at various stages throughout the front-end and back-end processes. For many manufacturers, it is not economically feasible or efficient to employ separate, redundant equipment for front-end/back-end materials and processes, solely for the sake of avoiding cross-contamination. Often, therefore, potentially harmful cross-contamination is mitigated by other means.
One approach to mitigating such cross-contamination involves inspecting representative samples, pulled from a production line immediately after processing or handling by a particular system or station (e.g., a wet clean system), for contaminants. If an unacceptable level of a contaminant is detected, the system or station in question is determined to be contamination and cleaned or decontaminated accordingly. Semiconductor wafers processed by the system or station in question, since the last satisfactory inspection, may then be individually examined or inspected for contamination, or scrapped.
There are, however, a number of issues that typically arise in conventional copper contamination inspection methods and systems. One issue concerns the frequency of inspections conducted. Commonly, representative samples are inspected on only a periodic basis. Inspections may occur once or twice per shift, or even once or twice per day. With a period of hours between inspections, a contaminated system or tool may have processed or handled a very large number of semiconductor wafers (e.g., >100). Once copper contamination has been detected in a particular system or tool, processing must stop to facilitate decontamination of the tool. Furthermore, each of the semiconductor wafers processed by that system or tool, since its last satisfactory inspection, must either be scrapped or individually inspected. Either way, time and money are wasted and lost.
Additionally, in many cases, conventional copper contamination inspection methods and systems require the removal of semiconductor wafers from the production line, to be taken to a separate contamination monitoring or detection system or apparatus. This introduces a certain degree of inefficiency into the production process. Many conventional inspection systems and methods are also rather time-consuming—often taking an hour or more to provide a complete analysis. Besides further increasing inefficiency of the process, a long inspection time allows for a large number of wafers to be processed by a potentially contaminated system or tool.
Another concern involves conventional contamination detection systems that rely on invasive contact with, or partial or complete dissolution of, the surface of the wafer being as tested. Where such detection systems are utilized, semiconductor wafers are therefore generally scrapped after inspection. Moreover, many conventional copper contamination detection systems and methods provide little, if any, indication of the specific position or source of copper contamination within a particular processing system or tool. Frequently, for example, contamination is determined by the amount of copper present in a given volume of dissolved semiconductor wafer.
As a result, there is a need for a versatile system for detecting and remediating cross-contamination, particularly copper cross-contamination, in semiconductor manufacturing processes. There is a further need for a versatile system that non-invasively detects and remediates cross-contamination—one that operates in-line and accurately indicates contamination origin(s)—providing timely, non-destructive remediation in an easy, efficient and cost-effective manner.